Download Homologous Recombination Generates T-Loop

Cell, Vol. 119, 355–368, October 29, 2004, Copyright 2004 by Cell Press
Homologous Recombination Generates
T-Loop-Sized Deletions at Human Telomeres
Richard C. Wang,1 Agata Smogorzewska,1,2
and Titia de Lange1,*
1
Laboratory for Cell Biology and Genetics
The Rockefeller University
1230 York Avenue
New York, New York 10021
Summary
The t-loop structure of mammalian telomeres is
thought to repress nonhomologous end joining (NHEJ)
at natural chromosome ends. Telomere NHEJ occurs
upon loss of TRF2, a telomeric protein implicated in
t-loop formation. Here we describe a mutant allele
of TRF2, TRF2⌬B, that suppressed NHEJ but induced
catastrophic deletions of telomeric DNA. The deletion
events were stochastic and occurred rapidly, generating dramatically shortened telomeres that were accompanied by a DNA damage response and induction
of senescence. TRF2⌬B-induced deletions depended
on XRCC3, a protein implicated in Holliday junction
resolution, and created t-loop-sized telomeric circles.
These telomeric circles were also detected in unperturbed cells and suggested that t-loop deletion by homologous recombination (HR) might contribute to
telomere attrition. Human ALT cells had abundant telomeric circles, pointing to frequent t-loop HR events
that could promote rolling circle replication of telomeres in the absence of telomerase. These findings
show that t-loop deletion by HR influences the integrity
and dynamics of mammalian telomeres.
Introduction
Changes in the length of human telomeres are relevant
to cancer and aging. Telomere attrition represents a
tumor suppressor pathway that limits the replicative potential of potential cancer cells. The gradual loss of telomeric DNA with each round of DNA replication depletes
the telomere reserve and leads to a growth arrest that is
accompanied by senescence or apoptosis. Most human
tumors need to bypass this proliferation block to reach
a clinically relevant mass. Telomerase activation has
surfaced as the main pathway for the escape from telomere-driven senescence or apoptosis, resulting in
nearly ubiquitous expression of telomerase during tumorigenesis (Shay and Bacchetti, 1997). Telomeraseindependent pathways (ALT) for telomere maintenance
are rare in most human cancers but significant in soft
tissue sarcomas (Henson et al., 2002). Telomere shortening has also been implicated in the age-related depletion of stem cell compartments and contributes to the
aging symptoms of dyskeratosis congenita patients
*Correspondence: [email protected]
2
Present address: Department of Pathology, Massachusetts General Hospital, 55 Fruit Street, WRN 219, Boston, Massachusetts
02114.
(Marciniak et al., 2000; Mitchell et al., 1999; Vulliamy et
al., 2001). Here, we document that HR can delete large
segments of telomeric DNA in human and mouse cells
and suggest that this pathway has implications for telomere-related disease states.
Telomeres protect chromosome ends from inappropriate DNA repair and prevent activation of DNA damage
checkpoints. Telomere protection is achieved through
telomerase-mediated maintenance of telomeric repeats, which endow chromosome ends with binding
sites for a protective protein complex. Mammalian telomeric TTAGGG repeats bind TRF2, a dimeric DNA binding protein with a key role in telomere protection (de
Lange, 2002). Dwindling presence of the TRF2 complex
on shortening telomeres is a likely cause of telomere
dysfunction (Karlseder et al., 2002). Like telomeres that
are depleted of TRF2, shortened telomeres become associated with DNA damage response factors (e.g.,
53BP1), forming cytological structures that are referred
to as telomere dysfunction induced foci (TIFs) (Bakkenist et al., 2004; d’Adda di Fagagna et al., 2003; Takai
et al., 2003). The ATM kinase plays a major role in the
telomere damage response and its ability to activate
p53 is largely responsible for telomere-driven cell cycle
arrest, senescence and/or apoptosis (d’Adda di Fagagna et al., 2003; Herbig et al., 2004; Karlseder et al.,
1999; Takai et al., 2003). TRF2 can function as a direct
inhibitor of the ATM kinase, and this feature can explain
why long telomeres, which contain 103–104 copies of
TRF2, do not activate the DNA damage pathway (Karlseder et al., 2004).
TRF2 is also crucial for the repression of NHEJ at
chromosome ends (Bailey et al., 2001; Smogorzewska
et al., 2002; van Steensel et al., 1998; Zhu et al., 2003).
Frequent telomere fusions occur when TRF2 is displaced from telomeres and in cells with excessive telomere erosion. The ability of TRF2 to protect telomeres
from NHEJ can be explained based on the formation of
t-loops (Stansel et al., 2001). T-loops sequester the 3⬘
end of the chromosome through strand invasion of the
telomeric overhang into the duplex TTAGGG repeat
array (Griffith et al., 1999). This telomeric configuration is
observed in purified telomeric DNA from human, mouse,
plant, and protozoan origin (Cesare et al., 2003; Griffith
et al., 1999; Munoz-Jordan et al., 2001; Murti and Prescott, 1999) and in native telomeric chromatin from
mouse and chicken cells (Nikitina and Woodcock, 2004).
When telomeres are in the t-loop configuration, the
NHEJ repair machinery may simply not have access to
chromosome ends, explaining the resistance of telomeres to fusion. However, t-loops resemble an intermediate in homologous recombination, potentially providing a substrate for their deletion through this pathway
(discussed in de Lange [2004], de Lange and Petrini
[2000], and Griffith et al. [1999]). Using a mutant allele
of TRF2, we show that HR can delete t-loop-sized telomeric segments resulting in rapid shortening of telomeres and generation of extrachromosomal telomeric
circles. In addition, we identify circular t-loop HR prod-
Cell
356
ucts in normal human cells and ALT cells, providing
evidence for the importance of HR in telomere dynamics.
Results
the duplex part of the telomere (van Steensel et al., 1998;
G. Celli and T.d.L., unpublished data).
The TRF2⌬B-induced reduction in mean telomere
length occurred in a variety of human cell lines (IMR90,
WI-38, HS68, BJ/hTERT, HT1080, and HeLa cells) (Figure 2C and data not shown). The change in the TTAGGG
repeat signals was quantified using equally loaded gels
and hybridization to chromosome internal control fragments to normalize the signals (see Supplemental Figure
S2 on the Cell web site). The quantitative analysis
showed that ⵑ15%–25% of the telomeric DNA was lost
in human and mouse cell lines infected with TRF2⌬B.
The loss of TTAGGG repeats was roughly proportional
to the length of the telomeres, such that longer telomeres lost larger segments of TTAGGG repeats. For
instance, mouse cells with telomeres in the 30–50 kb
range lost approximately the same fraction of their
TTAGGG repeats as did human cells with telomeres in
the 5–10 kb range (Figures 2C and D). The loss of telomeric DNA was rapid and occurred at a rate that far
exceeds the rate of telomere attrition in telomerasedeficient cells.
TRF2⌬B: Induction of TIFs and Senescence
without Telomere Fusions
TRF2 has a C-terminal Myb-type DNA domain and a
TRFH domain that mediates dimerization (Broccoli et
al., 1997; Fairall et al., 2001). These domains are very
similar to those of TRF1 and Taz1, the TRF1/2 ortholog
of Schizosaccharomyces pombe. TRF2 is unique in its
N terminus, which is comprised of a short segment of
basic amino acids. Deletion of this basic domain of TRF2
(TRF2⌬B) does not impede the DNA binding activity of
TRF2 (M. van Breughel and T.d.L., unpublished data) or
abrogate its ability to promote t-loop formation in vitro
(R. Stansel, T.d.L., and J.W. Griffith, unpublished data).
TRF2⌬B localizes to telomeres in vivo (van Steensel et
al., 1998) and can bind hRap1 (Li et al., 2000), and overexpression of this allele does not diminish the presence
of the Mre11 complex or ERCC1/XPF in the TRF2 complex (Zhu et al., 2000, 2003). In agreement with these
findings, telomeric ChIP on cells expressing TRF2⌬B revealed that the telomeres retained TRF2, TRF1, hRap1,
and Mre11 (see Supplemental Figure S1 at http://www.
cell.com/cgi/content/full/119/3/355/DC1/).
Despite its apparent proficiency as a telomere binding
protein, TRF2⌬B induced a cell cycle arrest in human
fibrosarcoma cells, suggesting that this allele created a
defect in telomere function (van Steensel et al., 1998).
In addition, TRF2⌬B resulted in senescence in primary
fibroblasts, as indicated by growth arrest, morphological changes, a gradual drop in BrdU incorporation, expression of senescence-associated (SA-) ␤-galactosidase (Dimri et al., 1995), and a pattern of p53, p16, p21,
and pRB expression typical of senescent cells (Figures
1A–1D). TRF2⌬B-expressing cells also developed 53BP1
foci, some of which colocalized with TRF1, a marker for
telomeres (Figure 1E). Such sites of colocalization of
DNA damage factors and telomeres (TIFs) are indicative
of dysfunctional telomeres. However, the TIFs were notably larger and less frequent than in cells expressing
the dominant-negative allele of TRF2 (TRF2⌬B⌬M) (Figure
1E). Furthermore, whereas TRF2⌬B⌬M results in telomere
fusions mediated by NHEJ, TRF2⌬B did not have this
phenotype (van Steensel et al. [1998] and see below),
suggesting that the telomere dysfunction induced by
TRF2⌬B was fundamentally distinct.
TRF2⌬B Induces Large Stochastic Postreplicative
Telomere Deletions
The loss of telomeric sequences was readily detectable
using telomere-specific fluorescence in situ hybridization (FISH). Metaphase spreads of mouse and human
fibroblasts showed an obvious overall loss in the
strength of telomeric FISH signals (Figures 3A and 3B).
The telomere deletions did not appear to affect all chromatids equally. While some telomeres appeared relatively unaffected, others were lost completely. Scoring
metaphases from several independent infections indicated that TRF2⌬B expression results in an ⵑ10-fold
increase in the number of chromatids lacking detectable
telomere signals in several human and mouse cell lines
(Supplemental Table S2). Although it is likely that signalfree ends still contain short stretches of telomeric DNA,
the data showed that the telomere signal loss is stochastic and due to large deletions in individual telomeres.
Metaphase spreads harvested within a few days after
introduction of TRF2⌬B also showed that telomere loss
was not random with regard to sister telomeres. Although many chromosomes lost more than one telomere, it was rare for both sister telomeres to lose their
signal (Figures 3A and 3B, Supplemental Table S2). This
implied that the deletions occurred after the replication
of telomeres.
TRF2⌬B-Induced Telomere Shortening
Analysis of genomic DNA showed that expression of
TRF2⌬B induced sudden shortening of the telomeric restriction fragments (Figure 2A). Inspection of denatured
telomeric restriction fragments on alkaline gels indicated that both the C-rich and the G-rich strands became shortened upon the expression of TRF2⌬B (Figure
2B). While the loss of telomeric DNA involved the duplex
part of the telomere, the G strand overhang signal did
not diminish or even increased (Figure 2A and data not
shown). This phenotype is clearly distinct from that of
TRF2⌬B⌬M or TRF2 gene deletion, which leads to a reduction of the G strand overhang signal but preservation of
Preferential Deletion of Telomeres Replicated
by Leading Strand DNA Synthesis
Inspection of the telomeric signal loss in TRF2⌬B-expressing
cells revealed that, on chromosomes with two deleted
telomeres, the remaining signals were very often diagonally positioned (Figure 3B). This finding suggested that
the telomere deletions might differentially affect telomeres generated by leading and lagging strand DNA
synthesis (referred to as leading strand and lagging
strand telomeres). Chromosome orientation (CO)-FISH
(Bailey et al., 1996) can be used to determine whether
a sister telomere is generated by leading or lagging
strand DNA synthesis (see Supplemental Figure S3A for
T-Loop Deletion by HR
357
Figure 1. Induction of Senescence and TIFs by TRF2⌬B
(A) Growth curves of IMR90 cells infected with the indicated retroviruses.
(B) Changes in BrdU incorporation of the cultures in (A). BrdU incorporation of vector control cells is set at 100% at day 0; the values represent
the mean of 4–5 experiments, and the SD is given. Method as described previously (Smogorzewska and de Lange, 2002).
(C) Immunoblot analysis of IMR90 cells expressing TRF2, TRF2⌬B, and vector control at day 14 of selection.
(D) TRF2⌬B-induced senescent morphology and SA-␤-gal expression. WI38 cells infected with the indicated virus stained for SA-␤-gal activity
at day 10 after selection.
(E) TIFs in BJ/hTERT infected for one day with the indicated viruses. DAPI (blue), 53BP1 IF (green), TRF1 IF (red). Enlarged images show
53BP1 foci at telomeres. Not all TRF2⌬B-induced 53BP1 foci coincided with detectable TRF1 signals.
schematic). Using CO-FISH on controls cells and cells
infected with TRF2⌬B, we usually found two diagonally
positioned lagging strand telomere signals per chromosome, indicating that the lagging strand telomere remained relatively intact (Supplemental Figure S3B). By
scoring the signals on 700 chromosomes from control
cells and TRF2⌬B cells, we found that the number of
lagging strand telomeres per chromosome remained
constant (2.1 in the vector control and 2.2 after TRF2⌬B
expression; n ⫽ 700 chromosomes for each group). In
contrast, CO-FISH to detect the leading strand telomeres showed dramatic signal loss after TRF2⌬B expression (Supplemental Figure S3C). Metaphase chromosomes often contained only one or no leading strand
CO-FISH signal. Overall, the leading strand signals per
chromosome was 0.74 after TRF2⌬B expression compared to 1.9 in the vector control (n ⫽ 650 chromosomes
for each group).
We next visualized both leading and lagging strand
telomeres on the same metaphase using CO-FISH (Figure 3C). Control metaphases possessed the expected
pattern of two different CO-FISH signals on each pair
of sister telomeres. However, metaphases from TRF2⌬Bexpressing cells showed a strong leading/lagging strand
signal imbalance with far fewer leading strand telomere
signals than expected. We conclude that TRF2⌬B expression results in a preferential deletion of leading strand
telomeres after telomere replication. Since sister asym-
Cell
358
Figure 2. TRF2⌬B Results in the Loss of Duplex Telomeric DNA but Not G Overhangs in
Human and Mouse Cells
(A) TRF2⌬B-induced shortening of the duplex
part of the telomere. BJ fibroblasts were harvested at the indicated time points after selection for infection with vector (⫺) or TRF2⌬B
retrovirus. Equal amounts of MboI- and AluIdigested genomic DNA were fractionated and
first probed with end-labeled (CCCTAA)4 under native conditions (left), followed by in situ
denaturation of the DNA in the gel and reprobing with the same probe (right). Position
and MWs (kb) of ␭ ⫻ HindIII fragments are
indicated to the left. Arrowheads indicate approximate median telomere restriction fragment lengths.
(B) Denaturing gel electrophoresis analysis of
C and G strand length after expression of
TRF2⌬B. BJ/hTERT:SV40LT fibroblasts were
harvested on day 10 of selection for infection
with vector (⫺) or TRF2⌬B (⫹). Equal amounts
of MboI- and AluI-digested genomic DNA
were denatured and fractionated on an alkaline gel. After neutralization and blotting, the
DNA was probed with labeled (CCCTAA)4 and,
after removal of signal, reprobed with (TTA
GGG)4. The marker is a self-ligated ladder of a
HindIII cut 2.6 kb plasmid containing TTAGGG
repeats (pTH5). MWs in kilobases (kb).
(C) TRF2⌬B-induced telomere shortening in
HeLa, BJ/hTERT, and IMR90. Telomere blots
of MboI- and AluI-digested genomic DNA
(from cells on day 4 of selection). Duplex
TTAGGG repeat signals were quantified as
shown in Supplemental Figure S2.
(D) TRF2⌬B-induced telomere shortening in
mouse cells. Genomic DNAs harvested 5
days after infection with vector (⫺) or TRF2⌬B
retrovirus (⫹), digested with HindIII, and separated on a CHEF gel. Telomeric signal were
quantified as in Supplemental Figure S2. The
band at 150 kb (*) presumably represents
telomeres with subtelomeric sequences lacking HindIII sites.
metry is erased after each round of DNA replication,
the preferential deletion of leading strand telomeres is
primarily observed in the first mitosis after TRF2⌬B has
its effect.
TRF2⌬B-Induced Deletions Depend
on XRCC3 and NBS1
Since the telomere deletions involved stochastic loss of
large segments of telomeric DNA from individual telomeres, we considered that they could be due to homologous recombination at the base of the t-loop. In order
to address the contribution of HR to telomere deletions,
we determined the effect of TRF2⌬B in cells with impaired
function of XRCC3, a RAD51 paralog that forms a complex with RAD51C (Liu et al., 2004). Although XRCC3
has been implicated in both early and late stages of
homologous recombination (Brenneman et al., 2002;
Pierce et al., 1999), recent studies have demonstrated
that XRCC3 and RAD51C are associated with Holliday
junction (HJ) resolvase activity in vitro (Liu et al., 2004).
Human HCT116 colon carcinoma cells in which both
copies of the XRCC3 gene were inactivated by gene
targeting (Yoshihara et al., 2004) were infected with
TRF2⌬B and examined for telomere deletions by quantitative genomic blotting (Figures 4A–4C). DNAs were digested with EcoRI/XbaI and first probed with a chromo-
T-Loop Deletion by HR
359
Figure 3. TRF2⌬B-Induced Telomere Deletions Are Stochastic and Preferentially Affect Leading Strand Telomeres
(A) Telomeric FISH on human BJ/hTERT:SV40LT metaphase chromosomes. Infected cells were harvested on day 6 after selection. Enlarged
images show chromosomes that lost two individual sister telomeres in an antiparallel fashion.
(B) Telomeric FISH on metaphase spreads from mouse NIH/3T3 cells. Infected cells were harvested 1 day after infection (no selection).
Enlarged images show chromosomes with unequal sister signals at both ends.
(C) Preferential deletion of leading strand telomeres detected by CO-FISH. Metaphases harvested from NIH/3T3 mouse cells incubated with
BrdU/BrdC were treated to remove the newly synthesized DNA strand. The remaining telomeric DNA strands were detected with TAMRA(TTAGGG)3 (red) and FITC-(CCCTAA)3 probe (green). The telomere replicated by leading strand synthesis is highlighted in red, and the telomere
replicated by lagging strand synthesis is highlighted in green. See Supplemental Figure S3 for CO-FISH schematic.
Cell
360
Figure 4. XRCC3 and NBS1 Are Required for TRF2⌬B-Induced Telomere Loss
(A) Requirement for XRCC3. Telomeric signals in human HCT116 cells, XRCC3⫺/⫺ HCT116 cells, and XRCC3⫺/⫺ HCT116 cells expressing
XRCC3 from exogenous cDNA infected with TRF2⌬B retrovirus or a vector control and harvested on day 5 after selection. EcoRI- and XbaIdigested DNA was probed for a chromosome-internal locus, quantified, and reprobed for telomeric repeats (shown), and telomere signal loss
was quantified as shown in Supplemental Figure S2. A subset of chromosome internal EcoRI and XbaI bands that crosshybridize with telomeric
probes are indicated with asterisks. The approximate positions of telomeric signals are indicated with gray boxes next to the lanes.
(B) Immunoblots for XRCC3, TRF2, and ␥-tubulin. XRCC3 (arrow) migrates just below a nonspecific band.
(C) Graph showing the effect of XRCC3 deficiency on telomere signal loss. Bars indicate mean loss of telomeric signals (⫾SD) induced by
TRF2⌬B from five independent genomic preps (three independent infections).
(D) Requirement for NBS1. Human NBS1-LB1 cells were complemented with stable expression of wt Nbs1, Nbs1 S343A, or vector control
using pLPC retroviruses. HTC75 fibrosarcoma cells, and NBS1-LB1 cells and their derivatives, were then infected with TRF2⌬B or the vector
control and harvested on day 5 after selection to prepare DNA and protein. Genomic DNA was digested with MboI and AluI, and telomeric
signals were quantified. The HTC75 and NBS1-LB1 cells were infected with TRF2⌬B or the vector control using pLPC; NBS1-LB1 derivative
lines were infected using pWZL.
(E) Immunoblots for Nbs1, TRF2, and ␥-tubulin.
(F) Graph showing the effect of NBS1 status on telomere signal loss. Bars indicate mean loss of telomeric signals (⫾SD) induced by TRF2⌬B
from four or more independent genomic preparations (four or more independent infections).
some internal histone gene probe to normalize for
loading and then reprobed with a telomeric probe (see
Supplemental Figure S2 for procedure). Triplicate experiments were used for the quantification of the effect of
XRCC3 status of TRF2⌬B-induced deletions. Although
the XRCC3⫺/⫺ cells expressed the same level of TRF2⌬B
as the control cells (Figure 4B), they did not show telomere deletions (Figures 4A and 4C). Expression of
XRCC3 from an introduced cDNA restored the telomere
deletion phenotype to the same level as the parental
HCT116 cells (Figures 4A–4C). Thus, TRF2⌬B-induced
telomere deletions required XRCC3, suggesting that HR
is involved in these events. A second RAD51 paralog,
RAD51D, has not been implicated in HJ resolution (Liu
et al., 2004) but is associated with telomeres (Tarsounas
et al., 2004). Using mouse cells lacking RAD51D due to
T-Loop Deletion by HR
361
gene targeting, we found that this RAD51 paralog was
not required for the TRF2⌬B-induced deletions (Supplemental Table S1).
A series of other genes involved in DNA repair (e.g.,
WRN helicase) and DNA damage signaling (e.g., ATM)
were tested for their contribution to TRF2⌬B-induced deletions (Supplemental Table S1). Among these, the only
gene defect that abrogated TRF2⌬B-mediated deletions
was a hypomorphic mutation in NBS1 (Figures 4D–4F).
The immortalized human NBS1-LB1 cell line lacks normal Nbs1 protein (Figure 4E) and also has a defect in
the nuclear localization of the other components of the
Mre11 complex (Lee et al., 2003). These cells showed
strongly diminished or no telomere deletion upon expression of TRF2⌬B. TRF2⌬B-induced deletions were also
abrogated in a primary human cell line deficient for Nbs1
(Supplemental Figure S4). Introduction of wt Nbs1 or an
Nbs1 signaling mutant (S343A; Lim et al., 2000) into
NBS1-LB1 cells restored the TRF2⌬B-induced telomere
deletions to the same level as HTC75 control cells. Thus,
Nbs1 and/or its partners in the Mre11 complex are required for telomere deletions after expression of TRF2⌬B.
HR at Telomeres Yields T-Loop-Sized
Circular DNAs
We analyzed the telomeric DNA of TRF2⌬B-expressing
cells using neutral-neutral 2D gel electrophoresis, which
separates telomeric restriction fragments first by size
and then by shape (Brewer and Fangman, 1987; Cohen
and Lavi, 1996). The 2D telomere blots of HeLa1.2.11
cells expressing TRF2⌬B showed the expected shortening of the telomeres compared to control cells (Figure
5). In addition, TRF2⌬B expression generated a new arc
of telomeric DNA, whose migration was consistent with
that of relaxed, double-stranded circles (Figure 5A). Expression of TRF2⌬B also induced the formation of this
novel arc in BJ/hTERT cells (Figure 5A) and mouse cells
(Supplemental Figure S5). These arcs were not induced
by overexpression of TRF1 or the dominant-negative
allele of TRF2, confirming their specificity for the TRF2⌬B
allele (Figure 5B and data not shown).
The circular structure of the telomeric molecules in the
TRF2⌬B-induced arc was confirmed by two approaches.
First, the arc comigrated with circularized ␭ ⫻ HindIII
DNA fragments (Figure 5B and Supplemental Figure S5).
Second, we reasoned that, if the arc represented circles
generated by HR within the telomere, they should be
devoid of subtelomeric sequences. To test this, genomic
DNA was digested with BglII and EcoRI, which results
in telomeric fragments carrying a subtelomeric repeat
element present at approximately 10% of human chromosome ends (de Lange et al., 1990). Probing 2D gels
with the subtelomeric probe did not reveal a circle arc
(Figure 5C), whereas reprobing the same gel with
TTAGGG repeats revealed the presence of the circle arc
as before. This result corroborated the view that the
circle arc represents deleted t-loops and also argued
against the possibility that the circle arc was due to DNA
replication or recombination intermediates that might
cause altered migration behavior in the second dimension.
The size distribution of the circular DNAs reflected the
size range of the telomeres. For instance, in HeLa1.2.11
cells with telomeres that are up to 25 kb in length, the
circle arc extended from 3 to 20 kb (Figure 5). By contrast, IMR90 cells with telomeres in the 6 to 10 kb range
yielded circles with a maximal size of only 9 kb (data
not shown). This size distribution of the circular DNAs
is consistent with the sizes of t-loops gauged from EM
analysis (Griffith et al., 1999). The EM data indicate that
t-loops have a broad size distribution in each individual
cell line but that the mean t-loop size correlates with
telomere lengths. Thus, the data are consistent with
deletion of t-loop-sized segments from individual telomeres. Due to the limitations of the resolution in the 2D
gels, circular products derived from very short telomeres
(⬍3 kb) were not detectable as separated circle arcs.
This prohibited the analysis of circular products in the
XRCC3 and Nbs1 mutant cell lines, which both have
very short TTAGGG repeat regions.
Long exposures of 2D gels of BJ/hTERT or HeLa experiments revealed the presence of low amounts of telomeric circles, even in vector control or uninfected cells
(Figure 5). The presence of the circle arc in unperturbed
cells indicated that circular telomeric DNA was not
strictly a result of TRF2⌬B expression but may also be
generated in small amounts by normal telomere metabolism. This finding is consistent with the occurrence of
occasional signal free ends in metaphase spreads of
control cells (Figure 3C). Previous studies of Hirt supernatant DNA from mammalian cells had also hinted at
the presence telomeric sequences among other circular
DNAs (Regev et al., 1998).
Telomeric Circles in ALT Cells
Budding yeast strains that lack telomerase can use one
of two HR pathways to maintain their telomeres (reviewed in Lundblad [2002]). HR has also been implicated
in the ALT pathway(s) that allow certain human cells to
maintain telomeres in the absence of telomerase
(Bechter et al., 2004; Dunham et al., 2000; LondonoVallejo et al., 2004). The findings with TRF2⌬B suggest
that HR is normally controlled at telomeres and predict
that the ALT phenotype would require derepression or
activation of telomeric HR. Therefore, we examined the
occurrence of telomeric circles in ALT cell lines by 2D
gel analysis (Figure 6). The results showed that five out
of six human ALT cell lines (VA13, Saos-2, SUSM-1, U2
OS, and MeT-4A) contained a strong circle arc, whereas
GM847 contained a circle arc of more moderate intensity. The circular telomeric DNA in the five ALT cell lines
appeared significantly more abundant than in telomerase-positive cells, reaching similar levels as in
cells expressing TRF2⌬B. The telomeric circles may explain the occasional detection of extrachromosomal
telomeric DNA in ALT cells (Ford et al., 2001; Tokutake
et al., 1998). Some of the ALT cell lines also had a prominent second arc that migrated at the position of supercoiled circular DNA (Figure 6), indicating covalently
closed circular DNAs that could be supercoiled due to
chromatinization.
Discussion
Human and mouse telomeres can occur in a t-loop configuration in which the 3⬘ single stranded overhang is
tucked into the double-stranded telomeric repeat region. Whereas the t-loop could protect telomeres from
Cell
362
Figure 5. TRF2⌬B-Induced Relaxed Telomeric Circles Detectable by 2D Gels
(A) 2D gel analysis of telomeric DNA from vector control and TRF2⌬B-expressing human cells. (Left) Schematic of the migration of linear dsDNA,
ssDNA, and relaxed dsDNA circles (Cohen and Lavi, 1996). DNA from HeLa1.2.11 (middle) and BJ/hTERT (right) cells, infected with the indicated
viruses separated by size (1D) and then shape (2D), blotted, and probed for telomeric DNA. MWs in kb. Arrows, telomeric circles induced by
TRF2⌬B. Insets, long exposures of part of the blots.
(B) Comigration of circularized ␭ ⫻ HindIII fragments with TRF2⌬B-induced telomeric circles. BJ/hTERT cells were infected with retroviruses
expressing TRF1 or TRF2⌬B. Genomic DNA was loaded adjacent to circularized ␭ ⫻ HindIII fragments. The blots were probed for the ␭ DNA
(middle), stripped, and reprobed for telomere repeats (left). The left and middle images were merged in Photoshop (right). Arrowheads, position
of circular DNAs. The 23 and 4.4 kb ␭ ⫻ HindIII fragments do not circularize due to their cos ends.
(C) Circular telomeric DNAs are not detected with a subtelomeric probe. (Left) Physical map of a subset of human telomeres that carry the
pTH2 subtelomeric repeat element (de Lange et al., 1990). Positions of the probe, subtelomeric repeats, telomeric repeats, and the BglII site
are shown. There is no EcoRI site distal from the BglII site. The indicated HeLa DNAs were digested with EcoRI and BglII and probed with
pTH2⌬ (middle), stripped, and reprobed with a TTAGGG repeat probe.
nonhomologous end joining, it resembles an intermediate in homologous recombination and could therefore
be at risk of inappropriate processing by this recombination pathway. The data presented here show that HR is
indeed a threat to human telomeres. Telomeres undergo
homologous recombination in cells that express a truncated form of the main protective factor at human telomeres, TRF2. Our data show that homologous recombination can delete large segments of telomeric DNA from
individual telomeres, generating circular telomeric DNAs
of t-loop size presumed to represent detached t-loops.
We refer to this process as t-loop HR to distinguish it
from homologous recombination between telomeres, a
process that could lead to exchange of telomeric DNA
but will not lead to net deletion of telomeric DNA. Because telomeric circles are also generated spontaneously in a variety of human cells, we propose that t-loop
HR contributes to stochastic shortening of human telomeres.
Telomere NHEJ and T-Loop HR: Double Jeopardy
at Chromosome Ends
NHEJ of telomeres is a detrimental event that creates
dicentric chromosomes and associated genome instability. At functional telomeres, NHEJ is repressed by
TRF2, presumably through formation of t-loops. How
telomeres prevent HR and what the consequences are
of this repair pathway at natural chromosome ends had
not been established. Our data indicate that TRF2 also
protects mammalian telomeres from potentially detrimental deletions through t-loop HR.
T-loop HR is enhanced upon introduction of TRF2⌬B,
a truncation mutant of TRF2 that lacks the N-terminal
basic domain. Unlike the dominant-negative allele of
T-Loop Deletion by HR
363
Figure 6. Prominent Telomeric Circle Arcs in ALT Cells
The schematic on the left shows the migration behavior of linear duplex and ssDNA, relaxed circles, and supercoiled circles. The 2D gels
show telomeric signals in MboI- and AluI-digested genomic DNA from the indicated human ALT lines. Five out of six lines (not GM847) had
elevated levels of circular telomeric DNA (black arrow) when compared to non-ALT cell lines. Four out of six lines (Saos-2, U-2 OS, MeT-4A,
and SUSM-1) possessed an additional arc of supercoiled circular telomeric DNA (gray arrow).
TRF2, TRF2⌬B localizes to telomeres, retains the ability
to protect telomeric overhangs from degradation, and
blocks NHEJ. However, TRF2⌬B induced stochastic homologous recombination events at individual telomeres,
resulting in deletions and t-loop-sized extrachromosomal circles. The simplest interpretation is that this
process of t-loop HR is normally repressed by TRF2
through a mechanism that involves its basic N-terminal
domain. Thus, TRF2⌬B appears to represent a dissociation of function mutant that represses NHEJ but unleashes HR when it displaces (part of) the endogenous
TRF2 from telomeres. It is anticipated that the basic
domain of TRF2 interacts with one or more other proteins involved in regulation of HR, but the relevant interacting partners are not known at this stage. The ability
of TRF2⌬B to promote HR appears to be specific to telomeres, as levels of sister chromatid exchange (SCE)
were comparable between vector control and TRF2⌬Bexpressing cells (data not shown).
T-loop HR is not observed upon the expression of
TRF2⌬B⌬M or when the TRF2 gene is deleted from mouse
cells, even when the NHEJ pathway is inactive (Smogorzewska et al., 2002; G. Celli and T.d.L., unpublished
data). This suggests that the binding of TRF2 to telomeres is required for efficient t-loop HR. Likely functions
for TRF2 in this regard are the formation of t-loops and
recruitment of the Mre11 complex, which is implicated in
t-loop HR. Both aspects of TRF2 function appear to be
unaffected by the deletion of the basic domain in TRF2⌬B.
It is not known whether TRF2 affects the efficiency of
other types of HR at telomeres (e.g., intertelomeric recombination events). A conserved role for duplex telomere repeat binding proteins in repressing HR at telomeres is consistent with the finding that S. pombe can
maintain its telomeres through a presumed recombination-based mechanism only when Taz1 is also deleted
(Nakamura et al., 1998).
Insight into the control of HR at mammalian telomeres
will be important for the understanding of telomere dynamics and may also reveal mechanisms by which HR
Cell
364
Figure 7. Speculative Model for T-Loop HR
(Top) Proposed structure of t-loops. (Middle) Branch migration at the strand invasion site of the telomere terminus results in the formation of
a Holliday junction. (Bottom) Two steps lead to t-loop deletion. Cleavage of the C strand at two positions by HJ resolvase (green arrowheads).
This process is proposed to require XRCC3. The second step involves nicking of the D loop by an unknown nuclease (open arrowhead). The
products are a shortened telomere and a relaxed telomeric circle. The shortened telomere might reform a small t-loop in a TRF2-dependent
manner (left) or, if the deletion is too extensive, might activate a DNA damage response and induce senescence. TRF2 is proposed to promote
t-loop formation, thereby creating the substrate for t-loop HR. The basic domain of TRF2 is proposed to suppress t-loop HR by inhibiting
branch migration and/or strand cleavage. In ALT cells, the telomeric circles resulting from t-loop HR could function as a template for rolling
circle replication and allow telomerase-independent telomere maintenance.
can be controlled. Our analysis of t-loop HR implicates
XRCC3, which, together with a second RAD51 paralog,
RAD51C, is associated with HJ resolvase activity in vitro
(Liu et al., 2004). T-loop HR also required a contribution
of the Nbs1 component of the Mre11 complex. The
Mre11 complex has been implicated in HR in yeast (Haber, 1998) and chicken DT40 cells (Tauchi et al., 2002).
The contribution of the Mre11 complex to mammalian
HR has been more difficult to study, since loss of this
complex is not compatible with cell viability (Luo et al.,
1999). Our results would suggest that Nbs1 and/or other
components of the Mre11 complex are required for
t-loop HR, but, given the multiplicity of functions ascribed to this complex (Haber, 1998), further analysis is
needed to establish the mechanism of its action in this
context. The signaling activity of the Mre11 complex is
unlikely to be required for t-loop HR, since the signaling-deficient Nbs1 S343A mutation does not abrogate
t-loop HR.
junction is a permanent feature of (some) T-loops. The
EM data on t-loops argue against extensive branch migration but do not exclude the presence of a small segment (⬍100 bp) of ds telomeric DNA at the t-loop base
(Griffith et al., 1999). The second step in t-loop HR is
HJ resolution by an activity that involves XRCC3 such
that the C strands are cleaved. In order to generate
extrachromosomal telomeric circles, a third step, cleavage of the D loop, would be required. The size distribution of telomeric circles formed by this pathway should
reflect the size distribution of t-loops, consistent with
the 2D gel data. The other product of the reaction is
the shortened telomere. This telomere will have a 3⬘
overhang and may be able to reform a t-loop, albeit a
smaller one. Depending on its length, the shortened
telomere may be fully functional but could be a substrate
for further t-loop HR, eventually leading to a critically
shortened telomere that activates the DNA damage response.
Model for T-Loop HR
We propose the following model for t-loop HR (Figure
7). We imagine that the first step is the branch migration
of the t-loop invasion site to yield a Holliday junction.
This step may not be necessary if the C strand terminus
normally invades the internal repeats and a Holliday
Yeast Telomere Rapid Deletions
and Mammalian T-Loop HR
The pioneering work of Lustig and colleagues has shown
that overelongated yeast telomeres can undergo rapid
deletions that reset telomeres to wild-type size (Bucholc
et al., 2001; Li and Lustig, 1996; reviewed in Lustig
T-Loop Deletion by HR
365
[2003]). These so-called TRD (telomere rapid deletion)
events have several features in common with the t-loop
HR reported here. TRD is due to intratelomeric homologous recombination. Physical mapping of deletions in
molecularly marked telomeres suggested a model in
which the deletions occur through a t-loop-like intermediate that is resolved by HR to yield the shortened telomere. The structure and fate of the deleted segment
has remained unknown, in part because yeast TRD is
rare (⬍10⫺3/telomere/division), prohibiting molecular
analysis of the immediate products.
While yeast TRD is enhanced in yKu70⌬ strains (Polotnianka et al., 1998), TRF2⌬B-induced t-loop HR was not
enhanced in mouse cells lacking DNA-PK components
or human cells treated with vanillin, a DNA-PKcs inhibitor (Supplemental Table S1). Further analysis will be
required to determine the possible role of DNA-PK in
mammalian spontaneous and TRF2⌬B-induced t-loop
HR. Another apparent difference is the requirement for
Nbs1. Human cells lacking Nbs1 fail to execute TRF2⌬Binduced t-loop HR, whereas TRD is only modestly affected by absence of Xrs2 (Bucholc et al., 2001), the
Nbs1-like protein of S. cerevisiae. However, deficiency
in the other components of the Mre11 complex, Mre11
and Rad50, abrogate TRD, stressing a parallel between
the mammalian and yeast pathways. Nbs1 mutations
in human cells may be more detrimental to the Mre11
complex than lack of Xrs2, perhaps because the nuclear
localization of the Mre11 complex is impeded when
Nbs1 is not fully functional. Collectively, the data support the view that yeast TRD is a good model for mammalian t-loop HR (Lustig, 2003). Yeast TRD counteracts
excessive elongation of telomeres and thus contributes
to the maintenance of normal telomere function. It is
not unlikely that HR at mammalian telomeres also serves
a function in normal telomere metabolism and that t-loop
HR only becomes a threat to telomeres when control
over this process is lost. This would be consistent with
the slight alterations in telomere function observed in
mouse cells deficient for RAD54 or RAD51D (Jaco et al.,
2003; Tarsounas et al., 2004).
Possible Contribution of T-Loop HR
to Telomere Attrition
The most detailed insight in telomere attrition derives
from the work of Kipling and colleagues, who examined
the length distribution of individual telomeres in clonal
telomerase-negative fibroblasts (Baird et al., 2003). Their
data show gradual progressive telomere erosion with
cell proliferation as well as stochastic changes in telomere length. The latter include telomere deletions with a
size range consistent with t-loop deletion. Such sudden
deletions can explain the kinetics of senescence in primary human cell strains (Rubelj and Vondracek, 1999).
Although telomere erosion is gradual and progressive,
affecting all telomeres to the same extent, entry into
senescence is not a synchronized event (Smith and
Whitney, 1980). Even in a clonal fibroblast line with telomeres of similar initial length, subclones are generated
that enter senescence earlier than the bulk of the culture.
Telomere length analysis in populations enriched for
such early senescent cells showed exaggerated telomere shortening (Martin-Ruiz et al., 2004). T-loop HR
could explain these observations. As cells progress
through their replicative lifespan, t-loop HR could generate individual cells in which critical telomere shortening
arises before general telomere erosion has taken its
effect in the bulk population. Thus, telomere-driven senescence would be due to a combination of t-loop HR
and gradual erosion; early senescent cells would be
more likely to suffer from the former; late senescent
cells would potentially suffer from both. The prediction
of this proposal is that cells deficient for t-loop HR
should still senesce but do so in a more synchronized
fashion than wild-type fibroblasts, and early senescent
subclones should not be formed.
Spontaneous Telomeric Circles in ALT Cells
The telomeric circles are abundant in most ALT cell lines,
consistent with the prior proposal that HR contributes
to telomere maintenance in these cells (Dunham et al.,
2000). Telomeric circles could be used as a template
for telomere extension by rolling circle replication, which
has been proposed as one of the possible mechanisms
for telomerase-independent telomere maintenance (de
Lange, 2004; Henson et al., 2002; Natarajan and McEachern, 2002). The predicted product of t-loop HR is a
relaxed circle with nicks in both strands (Figure 7), and
relaxed circles are observed in 2D gels. Nicked circles
will only allow a single round of rolling circle replication,
limiting the potential elongation of telomeres, but their
sealing by ligation would allow multiple rounds of replication. The supercoiled telomeric circle arcs we observe
in some ALT cells may be indicative of this process.
If the ALT pathway enhances HR at telomeres, telomere deletions are likely to be frequent in these cells.
This is consistent with early observations on an ALT cell
line that showed stochastic telomere deletions at low
frequency (Murnane et al., 1994) and also explains why
ALT cell lines often contain chromosome ends lacking
telomeric signals (Perrem et al., 2001). Critically shortened telomeres formed by t-loop deletion could affect
the growth characteristics of the cells. In glioblastoma
multiforme, the ALT pathway is associated with a strikingly better prognosis, suggesting a significant growth
difference between telomerase-positive tumor cells and
those that use ALT (Hakin-Smith et al., 2003). Furthermore, ALT cells have been reported to grow better when
telomerase is expressed (Stewart et al., 2002), a result
expected if telomerase helps to heal telomeres that have
been truncated by HR.
An Additional Function for Human Telomerase:
Counteracting Catastrophic Telomere Deletions
Telomerase is generally thought to counteract the gradual progressive erosion of telomeres, which requires
synthesis of up to 200 nt of telomeric DNA per chromosome end per cell division. Counteracting the large deletions generated by t-loop HR is more tasking, requiring
the synthesis of up to several kb of telomeric DNA at
the affected ends while the other telomeres remain at
roughly the same size. This process could explain reports of telomere length-independent functions of telomerase. Telomerase expression has been noted to
improve the growth characteristics and telomere function in settings where the enzyme had no detectable
Cell
366
effect on the mean telomere length (e.g., Stewart et al.,
2002; Zhu et al., 1999; reviewed in Blackburn [2001]).
These findings have been interpreted as evidence for a
“capping” function of telomerase. We propose that, in
at least some of these experiments, telomerase is required for cell survival because it can repair telomeres
deleted by t-loop HR.
A possible role for telomerase in counteracting t-loop
HR has implications for telomerase-based cancer therapy. Although telomerase inhibition kills cancer cells,
the efficacy of telomerase inhibitors has been debated
because of their predicted phenotypic lag (Shay and
Wright, 2002). This lag would represent the proliferation
needed to deplete the telomere reserve. However, the
effect of telomerase inhibition can be remarkably fast
in vitro (Hahn et al., 1999). Occasional large telomere
deletions through HR can now explain the rapid effect
of telomerase inhibition. Cells in which one or more
telomeres have undergone t-loop HR and induce a DNA
damage response will not proliferate unless telomerase
resynthesizes enough telomeric DNA to reconstitute
telomere function. Thus, telomerase inhibition may be
a fast-acting antitumor strategy in cancers with frequent
t-loop HR, and conditions that stimulate t-loop HR could
potentially enhance the efficacy of telomerase therapy
in the clinic.
immunoblotting, ChIP, FISH, and IF procedures are described in
Supplemental Data.
Experimental Procedures
Baird, D.M., Rowson, J., Wynford-Thomas, D., and Kipling, D. (2003).
Extensive allelic variation and ultrashort telomeres in senescent
human cells. Nat. Genet. 33, 203–207.
2D Gel Electrophoresis and Genomic Blotting
Isolation and digestion of genomic DNA were performed according
to standard protocols (Karlseder et al., 2002). For standard telomere
blots, MboI- and AluI-digested genomic DNA was fractionated on
a 0.7% agarose gel containing 0.1 ␮g/ml ethidium bromide (EtBr) in
1 ⫻ TAE at ⵑ2V/cm overnight. Neutral-neutral 2D gel electrophoresis
was performed according to the protocols established by Brewer
and Fangman (http://fangman-brewer.genetics.washington.edu/
2Dgel.html/) with the modifications described by Cohen and Lavi
(1996).
MEF Preparations and Pulsed-Field Gel Electrophoresis
MEFs were trypsinized, washed once in PBS, and embedded in 1%
agarose plugs at a concentration of 1.5 ⫻ 106 cells per 100 ␮l plug.
Plugs were incubated in proteinase K digestion buffer (10 mM Tris
[pH 7.9], 200 mM EDTA [pH 8.0], 0.2% sodium deoxycholate, 1%
sodium lauryl sarcosine, and 500 mg/ml proteinase K; 1 ml per plug)
at 50⬚C overnight (o/n). The plugs were washed extensively in TE
(10 mM Tris-HCl [pH 7.5], 1 mM EDTA), equilibrated in digestion
buffer for 1 hr, and digested with 120 U of HindIII and 10 ␮g RNase
A in 1 ml for 12 hr. After digestion, plugs were washed in TE and
equilibrated in 0.5 ⫻ TBE. The plugs were fractionated on a CHEFDRII PFGE (Bio-Rad) in a 1% agarose gel in 0.5⫻ TBE at 6 V/cm
for 20 hr at 14⬚C.
Measurements of Reduction in Telomeric Repeat Signals
Genomic DNA was prepared from cells infected in parallel with
vector or TRF2⌬B retroviruses. After digestion and quantification by
Hoechst fluorometry, equal amounts of DNA (ⵑ4 ␮g) were fractionated, and equal loading was confirmed by EtBr staining. Blots were
generally first probed and quantitated for a chromosome internal
loading control (see Supplemental Data). The blots were then
stripped with boiling 0.1% SDS, rinsed in 2 ⫻ SSC, and hybridized
with a TTAGGG repeat probe (pSP73.Sty11; de Lange [1992]), and
the telomere signal was quantified in ImageQuant. The telomeric
signal was normalized first to the loading control for all lanes; the
normalized values for vector- and TRF2⌬B-infected samples were
compared and expressed as % signal change in the TRF2⌬B
samples.
Additional technical details and cell culture, viral gene delivery,
Acknowledgments
We are grateful to Kiyoshi Miyagawa, Michael Liskay, John Petrini,
Thanos Halazonetis, Paul Hasty, Lexicon Genetics, Fred Alt,
Margeret Zdzienicka, Thomas Kunkel, Alan Clark, Winfried Edelman,
Hein te Riele, Phillip Smiraldo, and Douglas Pittman for generous
gifts of reagents and cell lines. We thank Xu-Dong Zhu, Giulia Celli,
Jan Karlseder, Hiro Takai, Diego Loayza, Dirk Hockemeyer, Susan
Bailey, and Maria Blasco for invaluable assistance with protocols.
Jim Haber and members of the de Lange lab are thanked for helpful
discussions and comments on the paper. This work was supported
by a grant to T.d.L. from the NIH (GM49046). R.C.W. and A.S. were
supported by an NIH MSTP grant (GM07739) to the Cornell/RU/MSK
Tri-Institutional MD/PhD program.
Received: July 21, 2004
Revised: September 16, 2004
Accepted: September 28, 2004
Published: October 28, 2004
References
Bailey, S.M., Goodwin, E.H., Meyne, J., and Cornforth, M.N. (1996).
CO-FISH reveals inversions associated with isochromosome formation. Mutagenesis 11, 139–144.
Bailey, S.M., Cornforth, M.N., Kurimasa, A., Chen, D.J., and Goodwin, E.H. (2001). Strand-specific postreplicative processing of mammalian telomeres. Science 293, 2462–2465.
Bakkenist, C.J., Drissi, R., Wu, J., Kastan, M.B., and Dome, J.S.
(2004). Disappearance of the telomere dysfunction-induced stress
response in fully senescent cells. Cancer Res. 64, 3748–3752.
Bechter, O.E., Shay, J.W., and Wright, W.E. (2004). The frequency
of homologous recombination in human ALT cells. Cell Cycle 3,
547–549. Published online May 10, 2004. PMID: 15034305
Blackburn, E.H. (2001). Switching and signaling at the telomere. Cell
106, 661–673.
Brenneman, M.A., Wagener, B.M., Miller, C.A., Allen, C., and Nickoloff, J.A. (2002). XRCC3 controls the fidelity of homologous recombination: roles for XRCC3 in late stages of recombination. Mol. Cell
10, 387–395.
Brewer, B.J., and Fangman, W.L. (1987). The localization of replication origins on ARS plasmids in S. cerevisiae. Cell 51, 463–471.
Broccoli, D., Smogorzewska, A., Chong, L., and de Lange, T. (1997).
Human telomeres contain two distinct Myb-related proteins, TRF1
and TRF2. Nat. Genet. 17, 231–235.
Bucholc, M., Park, Y., and Lustig, A.J. (2001). Intrachromatid excision of telomeric DNA as a mechanism for telomere size control in
Saccharomyces cerevisiae. Mol. Cell. Biol. 21, 6559–6573.
Cesare, A.J., Quinney, N., Willcox, S., Subramanian, D., and Griffith,
J.D. (2003). Telomere looping in P. sativum (common garden pea).
Plant J. 36, 271–279.
Cohen, S., and Lavi, S. (1996). Induction of circles of heterogeneous
sizes in carcinogen-treated cells: two-dimensional gel analysis of
circular DNA molecules. Mol. Cell. Biol. 16, 2002–2014.
d’Adda di Fagagna, F., Reaper, P.M., Clay-Farrace, L., Fiegler, H.,
Carr, P., Von Zglinicki, T., Saretzki, G., Carter, N.P., and Jackson,
S.P. (2003). A DNA damage checkpoint response in telomere-initiated senescence. Nature 426, 194–198.
de Lange, T. (1992). Human telomeres are attached to the nuclear
matrix. EMBO J. 11, 717–724.
de Lange, T. (2002). Protection of mammalian telomeres. Oncogene
21, 532–540.
de Lange, T. (2004). t-loops and the origin of telomeres. Nat. Rev.
Mol. Cell Biol. 5, 323–329.
T-Loop Deletion by HR
367
de Lange, T., and Petrini, J. (2000). A new connection at human
telomeres: association of the Mre11 complex with TRF2. Cold Spring
Harb. Symp. Quant. Biol. 65, 265–273.
de Lange, T., Shiue, L., Myers, R.M., Cox, D.R., Naylor, S.L., Killery,
A.M., and Varmus, H.E. (1990). Structure and variability of human
chromosome ends. Mol. Cell. Biol. 10, 518–527.
Dimri, G.P., Lee, X., Basile, G., Acosta, M., Scott, G., Roskelley, C.,
Medrano, E.E., Linskens, M., Rubelj, I., Pereira-Smith, O., et al.
(1995). A biomarker that identifies senescent human cells in culture
and in aging skin in vivo. Proc. Natl. Acad. Sci. USA 92, 9363–9367.
characterized by high rates of telomeric exchange. Cancer Res.
64, 2324–2327.
Lundblad, V. (2002). Telomere maintenance without telomerase. Oncogene 21, 522–531.
Luo, G., Yao, M.S., Bender, C.F., Mills, M., Bladl, A.R., Bradley, A.,
and Petrini, J.H. (1999). Disruption of mRad50 causes embryonic
stem cell lethality, abnormal embryonic development, and sensitivity
to ionizing radiation. Proc. Natl. Acad. Sci. USA 96, 7376–7381.
Lustig, A.J. (2003). Clues to catastrophic telomere loss in mammals
from yeast telomere rapid deletion. Nat. Rev. Genet. 4, 916–923.
Dunham, M.A., Neumann, A.A., Fasching, C.L., and Reddel, R.R.
(2000). Telomere maintenance by recombination in human cells. Nat.
Genet. 26, 447–450.
Marciniak, R.A., Johnson, F.B., and Guarente, L. (2000). Dyskeratosis congenita, telomeres and human ageing. Trends Genet. 16,
193–195.
Fairall, L., Chapman, L., Moss, H., de Lange, T., and Rhodes, D.
(2001). Structure of the TRFH dimerization domain of the human
telomeric proteins TRF1 and TRF2. Mol. Cell 8, 351–361.
Martin-Ruiz, C., Saretzki, G., Petrie, J., Ladhoff, J., Jeyapalan, J.,
Wei, W., Sedivy, J., and Von Zglinicki, T. (2004). Stochastic variation
in telomere shortening rate causes heterogeneity of human fibroblast replicative life span. J. Biol. Chem. 279, 17826–17833.
Ford, L.P., Zou, Y., Pongracz, K., Gryaznov, S.M., Shay, J.W., and
Wright, W.E. (2001). Telomerase can inhibit the recombinationbased pathway of telomere maintenance in human cells. J. Biol.
Chem. 276, 32198–32203.
Griffith, J.D., Comeau, L., Rosenfield, S., Stansel, R.M., Bianchi, A.,
Moss, H., and de Lange, T. (1999). Mammalian telomeres end in a
large duplex loop. Cell 97, 503–514.
Haber, J.E. (1998). The many interfaces of Mre11. Cell 95, 583–586.
Hahn, W.C., Stewart, S.A., Brooks, M.W., York, S.G., Eaton, E., Kurachi, A., Beijersbergen, R.L., Knoll, J.H., Meyerson, M., and Weinberg,
R.A. (1999). Inhibition of telomerase limits the growth of human
cancer cells. Nat. Med. 5, 1164–1170.
Hakin-Smith, V., Jellinek, D.A., Levy, D., Carroll, T., Teo, M., Timperley, W.R., McKay, M.J., Reddel, R.R., and Royds, J.A. (2003). Alternative lengthening of telomeres and survival in patients with glioblastoma multiforme. Lancet 361, 836–838.
Henson, J.D., Neumann, A.A., Yeager, T.R., and Reddel, R.R. (2002).
Alternative lengthening of telomeres in mammalian cells. Oncogene
21, 598–610.
Herbig, U., Jobling, W.A., Chen, B.P., Chen, D.J., and Sedivy, J.M.
(2004). Telomere shortening triggers senescence of human cells
through a pathway involving ATM, p53, and p21(CIP1), but not
p16(INK4a). Mol. Cell 14, 501–513.
Jaco, I., Munoz, P., Goytisolo, F., Wesoly, J., Bailey, S., Taccioli, G.,
and Blasco, M.A. (2003). Role of mammalian Rad54 in telomere
length maintenance. Mol. Cell. Biol. 23, 5572–5580.
Karlseder, J., Broccoli, D., Dai, Y., Hardy, S., and de Lange, T. (1999).
p53- and ATM-dependent apoptosis induced by telomeres lacking
TRF2. Science 283, 1321–1325.
Karlseder, J., Smogorzewska, A., and de Lange, T. (2002). Senescence induced by altered telomere state, not telomere loss. Science
295, 2446–2449.
Karlseder, J., Hoke, K., Mirzoeva, O.K., Bakkenist, C., Kastan, M.B.,
Petrini, J.H., and Lange Td, T. (2004). The telomeric protein TRF2
binds the ATM kinase and can inhibit the ATM-dependent DNA
damage response. PLoS Biol 2(8): e240 DOI: 10.1371/journal.pbio.
0020240.
Lee, J.H., Xu, B., Lee, C.H., Ahn, J.Y., Song, M.S., Lee, H., Canman,
C.E., Lee, J.S., Kastan, M.B., and Lim, D.S. (2003). Distinct functions
of Nijmegen breakage syndrome in ataxia telangiectasia mutateddependent responses to DNA damage. Mol. Cancer Res. 1, 674–681.
Li, B., and Lustig, A.J. (1996). A novel mechanism for telomere size
control in Saccharomyces cerevisiae. Genes Dev. 10, 1310–1326.
Li, B., Oestreich, S., and de Lange, T. (2000). Identification of human
Rap1: implications for telomere evolution. Cell 101, 471–483.
Lim, D.S., Kim, S.T., Xu, B., Maser, R.S., Lin, J., Petrini, J.H., and
Kastan, M.B. (2000). ATM phosphorylates p95/nbs1 in an S-phase
checkpoint pathway. Nature 404, 613–617.
Liu, Y., Masson, J.Y., Shah, R., O’Regan, P., and West, S.C. (2004).
RAD51C is required for Holliday junction processing in mammalian
cells. Science 303, 243–246.
Londono-Vallejo, J.A., Der-Sarkissian, H., Cazes, L., Bacchetti, S.,
and Reddel, R.R. (2004). Alternative lengthening of telomeres is
Mitchell, J.R., Wood, E., and Collins, K. (1999). A telomerase component is defective in the human disease dyskeratosis congenita. Nature 402, 551–555.
Muñoz-Jordan, J.L., Cross, G.A., de Lange, T., and Griffith, J.D.
(2001). t-loops at trypanosome telomeres. EMBO J. 20, 579–588.
Murnane, J.P., Sabatier, L., Marder, B.A., and Morgan, W.F. (1994).
Telomere dynamics in an immortal human cell line. EMBO J. 13,
4953–4962.
Murti, K.G., and Prescott, D.M. (1999). Telomeres of polytene chromosomes in a ciliated protozoan terminate in duplex DNA loops.
Proc. Natl. Acad. Sci. USA 96, 14436–14439.
Nakamura, T.M., Cooper, J.P., and Cech, T.R. (1998). Two modes of
survival of fission yeast without telomerase. Science 282, 493–496.
Natarajan, S., and McEachern, M.J. (2002). Recombinational telomere elongation promoted by DNA circles. Mol. Cell. Biol. 22, 4512–
4521.
Nikitina, T., and Woodcock, C.L. (2004). Closed chromatin loops at
the ends of chromosomes. J. Cell Biol. 166, 161–165. Published
online July 12, 2004. 10.1083/jcb.200403118
Perrem, K., Colgin, L.M., Neumann, A.A., Yeager, T.R., and Reddel,
R.R. (2001). Coexistence of alternative lengthening of telomeres
and telomerase in hTERT-transfected GM847 cells. Mol. Cell. Biol.
21, 3862–3875.
Pierce, A.J., Johnson, R.D., Thompson, L.H., and Jasin, M. (1999).
XRCC3 promotes homology-directed repair of DNA damage in mammalian cells. Genes Dev. 13, 2633–2638.
Polotnianka, R.M., Li, J., and Lustig, A.J. (1998). The yeast Ku heterodimer is essential for protection of the telomere against nucleolytic
and recombinational activities. Curr. Biol. 8, 831–834.
Regev, A., Cohen, S., Cohen, E., Bar-Am, I., and Lavi, S. (1998).
Telomeric repeats on small polydisperse circular DNA (spcDNA) and
genomic instability. Oncogene 17, 3455–3461.
Rubelj, I., and Vondracek, Z. (1999). Stochastic mechanism of cellular aging—abrupt telomere shortening as a model for stochastic
nature of cellular aging. J. Theor. Biol. 197, 425–438.
Shay, J.W., and Bacchetti, S. (1997). A survey of telomerase activity
in human cancer. Eur. J. Cancer 33, 787–791.
Shay, J.W., and Wright, W.E. (2002). Telomerase: a target for cancer
therapeutics. Cancer Cell 2, 257–265.
Smith, J.R., and Whitney, R.G. (1980). Intraclonal variation in proliferative potential of human diploid fibroblasts: stochastic mechanism
for cellular aging. Science 207, 82–84.
Smogorzewska, A., and de Lange, T. (2002). Different telomere damage signaling pathways in human and mouse cells. EMBO J. 21,
4338–4348.
Smogorzewska, A., Karlseder, J., Holtgreve-Grez, H., Jauch, A., and
de Lange, T. (2002). DNA ligase IV-dependent NHEJ of deprotected
mammalian telomeres in G1 and G2. Curr. Biol. 12, 1635–1644.
Stansel, R.M., de Lange, T., and Griffith, J.D. (2001). T-loop assembly
in vitro involves binding of TRF2 near the 3⬘ telomeric overhang.
EMBO J. 20, E5532–E5540.
Cell
368
Stewart, S.A., Hahn, W.C., O’Connor, B.F., Banner, E.N., Lundberg,
A.S., Modha, P., Mizuno, H., Brooks, M.W., Fleming, M., Zimonjic,
D.B., et al. (2002). Telomerase contributes to tumorigenesis by a
telomere length-independent mechanism. Proc. Natl. Acad. Sci.
USA 99, 12606–12611. Published online August 22, 2002.
10.1073/pnas.182407599
Takai, H., Smogorzewska, A., and de Lange, T. (2003). DNA damage
foci at dysfunctional telomeres. Curr. Biol. 13, 1549–1556.
Tarsounas, M., Munoz, P., Claas, A., Smiraldo, P.G., Pittman, D.L.,
Blasco, M.A., and West, S.C. (2004). Telomere maintenance requires
the RAD51D recombination/repair protein. Cell 117, 337–347.
Tauchi, H., Kobayashi, J., Morishima, K., van Gent, D.C., Shiraishi,
T., Verkaik, N.S., vanHeems, D., Ito, E., Nakamura, A., Sonoda, E.,
et al. (2002). Nbs1 is essential for DNA repair by homologous recombination in higher vertebrate cells. Nature 420, 93–98.
Tokutake, Y., Matsumoto, T., Watanabe, T., Maeda, S., Tahara, H.,
Sakamoto, S., Niida, H., Sugimoto, M., Ide, T., and Furuichi, Y. (1998).
Extra-chromosomal telomere repeat DNA in telomerase-negative
immortalized cell lines. Biochem. Biophys. Res. Commun. 247,
765–772.
van Steensel, B., Smogorzewska, A., and de Lange, T. (1998). TRF2
protects human telomeres from end-to-end fusions. Cell 92,
401–413.
Vulliamy, T., Marrone, A., Goldman, F., Dearlove, A., Bessler, M.,
Mason, P.J., and Dokal, I. (2001). The RNA component of telomerase
is mutated in autosomal dominant dyskeratosis congenita. Nature
413, 432–435.
Yoshihara, T., Ishida, M., Kinomura, A., Katsura, M., Tsuruga, T.,
Tashiro, S., Asahara, T., and Miyagawa, K. (2004). XRCC3 deficiency
results in a defect in recombination and increased endoreduplication in human cells. EMBO J. 23, 670–680.
Zhu, J., Wang, H., Bishop, J.M., and Blackburn, E.H. (1999). Telomerase extends the lifespan of virus-transformed human cells
without net telomere lengthening. Proc. Natl. Acad. Sci. USA 96,
3723–3728.
Zhu, X.D., Kuster, B., Mann, M., Petrini, J.H., and de Lange, T. (2000).
Cell-cycle-regulated association of RAD50/MRE11/NBS1 with TRF2
and human telomeres. Nat. Genet. 25, 347–352.
Zhu, X.D., Niedernhofer, L., Kuster, B., Mann, M., Hoeijmakers, J.H.,
and de Lange, T. (2003). ERCC1/XPF Removes the 3⬘ overhang from
uncapped telomeres and represses formation of telomeric DNAcontaining double minute chromosomes. Mol. Cell 12, 1489–1498.
Note Added in Proof
Cesare and Griffith independently observed circular molecules in
ALT cells by electron microscopic analysis of purified telomeric
DNA: Cesare, A.J., and Griffith, J.D. (2004). Telomeric DNA in ALT
cells is characterized by free telomeric circles and heterogeneous
t-loops. Mol. Cell Biol., in press.
Supplemental Data for
Wang et al., Cell 119, pp. 355–368
Supplemental Experimental Procedures
Cell Culture
WI-38 (ATCC), IMR90 (ATCC), GM07166B (Nbs1) (Coriell), AG03141C (WRN) (ATCC),
AG04405A (ATM) (ATCC), and AG02496 (ATM) (ATCC) were grown in DMEM with 100 U of
penicillin and 0.1 mg of streptomycin, 2 mM L-glutamine, 0.1 mM nonessential amino acids
(GIBCO), 1 mM sodium pyruvate (Sigma) and 15% FBS. U-2 OS (ATCC) and Saos-2 (ATCC)
were grown in McCoy’s 5a medium with supplements and 10% or 15% FBS, respectively. MeT4A cells were grown in RPMI with 10% FBS. SUSM-1, GM847, VA13, HTC75, HeLaI.2.11,
NBS1-LB1 (Kraakman-van der Zwet et al. [1999]; a gift from M. Zdzienicka), Phoenix
amphotropic, and Phoenix ecotropic packaging cells were grown in DMEM with supplements
and 10% FBS. BJ/hTERT cells (Clontech) were grown according to the manufacturer’s
specifications. NIH3T3, ERCC1-/- and littermate cells (Zhu et al., 2003), DNA-PKcs-/- (Gao et
al. [1998]; a generous gift from F. Alt), Ku86-/-p53-/- and Ku86+/+p53-/- (Zhu et al., 1996), and
RAD51-/-p53-/- and RAD51+/+p53-/- (Tarsounas et al. [2004]; a generous gift from D. Pittman)
were grown in DMEM with supplements and 10% FBS (or 15% heat inactivated FBS and 50 M
2-b-mercaptoethanol for primary MEFs). Growth curves, BrdU labeling, and SA-b-galactosidase
assays were done as previously described (Smogorzewska and de Lange, 2002). TIF analysis
on BJ/hTERT cells was performed as previously described (Takai et al., 2003).
Retroviral Gene Delivery
Either pLPCpuro or pWZLhygro was used as retroviral vectors for expression of TRF2,
TRF2B, and TRF1. Primary MEFs were transformed with pBabeNeoSV40largeT at early
passage. Retroviral gene delivery was performed as described (Karlseder et al., 2002) except
that cells were infected three times at 12 hr intervals. Primary AT fibroblasts (AG04405A,
AG02496) were infected three times at 24 hr intervals. Retrovirally infected cells were selected
in the appropriate drugs (puromycin 2 g/ml, hygromycin 90 g/ml, or G418 600 g/ml) for 4–5
days until uninfected control cells were completed selected. ERCC1+/+ and ERCC1-/- cells
were selected in 3 g/ml puromycin. Infection of NIH/3T3 cells proved to be very efficient (>90%
expression after three rounds). Thus, for better preservation of telomere signals, FISH was
performed on NIH/3T3 cells immediately after a 24 hr recovery from two 12 hr infections without
selection.
Immunoblotting and Chromatin Immunoprecipitation
Whole-cell lysates were prepared and fractionated as previously described (Smogorzewska and
de Lange [2002] and Loayza and De Lange [2003]). For chromatin immunoprecipitation (ChIP),
25 l of the following crude sera were used: TRF1 (371), TRF2 (647), TRF2 basic domain (508),
hRap1 (765), Mre11 (874), Mre11 preimmune (874). For Westerns, the following antibodies
were used at the manufacturers’ recommended dilutions (except where noted): p53 (DO-1), p21
(Upstate F-5), Rb (Pharmingen 14001A), p16 (Novacastra NCL p16), cyclin A (Santa Cruz sc239), cyclin D (Santa Cruz), Nbs1 (generous gift from J. Petrini, Hp95 #16/9, 1:20,000), XRCC3
(Chemicon 3776, 5g/ml), g-tubulin (Sigma GTU88, 1:20,000).
Measurements of Reduction in Telomeric Repeat Signals
Cells were infected with vector or TRF2B retroviruses and processed in parallel. Cells were
harvested, and genomic DNA was isolated, digested with the indicated restriction enzymes, and
quantified by Hoechst flourometry after digestion. Equal amounts of DNA (4 mg) were
fractionated, and equal loading was confirmed by EtBr staining. Blots were generally first probed
for a loading control. For human DNA, two control probes were used—a 475 bp Histone H1.3
(PCR primer sequences available upon request) or a fraction of genomic sequences that
migrate at ~23 kb after MboI + AluI digestion (referred to as genomic DNA in Supplemental
Figure S2) that contains unknown repetitive sequences lacking MboI/AluI sites. The blots were
exposed on PhophorImager screens, and the loading control signal was quantified using
Imagequant. For mouse DNA, the mTRF2 probe was used as described above, and bands
representing the mTRF2 gene were used as loading control. The blots were then stripped with
boiling 0.1%SDS, rinsed in 2xSSC, and hybridized with a TTAGGG repeat probe (pSP73.Sty11;
de Lange [1992]), and the telomere signal was quantified in Imagequant. The telomeric signal
was normalized to the loading control for all lanes, and the normalized values for vector and
TRF2B infected samples were compared and expressed as percent signal change in the
TRF2B samples.
2D Gel Electrophoresis and Genomic Blotting
Isolation and digestion of genomic DNA were performed according to standard protocols
(Karlseder et al., 2002). For standard telomere blots, MboI- and AluI-digested genomic DNA
was fractionated on a 0.7% agarose gel containing 0.1 g/ml ethidium bromide (EtBr) in 1 ×
TAE at ~2 V/cm overnight. Neutral-neutral 2D gel electrophoresis was performed according to
the protocols established by Brewer and Fangman (http://fangmanbrewer.genetics.washington.edu/2Dgel.html) with the modifications described by Cohen and
Lavi (1996). An equal amount (10–15 g) of digested genomic DNA was separated on a 0.4%
agarose gel in 1 × TBE at 1 V/cm for 20–24 hr. The gel was stained with 0.3 g EtBr per ml,
and lanes were cut, placed orthogonally, and cast in 1.1% agarose in 1 × TBE containing 0.3 g
EtBr per ml. The second dimension was run in at 4–5 V/cm for 3.5 hr at RT. Genomic blotting
was performed as previously described (van Steensel and de Lange, 1997). Circularized marker
DNA was generated by ligating HindIII-digested I DNA (NEB) at 5 ng/l overnight at 16°C. The
ligated DNA was ethanol precipitated in the presence of glycogen and resuspended at 500
ng/l; 500 ng was used per lane. To generate the ladder probe, HindIII cut I DNA was digested
with BstEII and gel purified. Approximately 2 g of the digested I DNA was labeled with T4 PNK
and g32P-ATP.
Metaphase Chromosome Spreads, FISH, and CO-FISH
Metaphase spreads were prepared as previously described (van Steensel et al., 1998). FISH
was performed using a FITC-conjugated PNA probe specific for the G strand (FITC-5¢[CCCTAA]3-3¢) (Applied Biosystems). CO-FISH was performed using either the FITC-TelC
probe or a TAMRA-conjugated PNA probe specific for the C strand (TAMRA-5-[TTAGGG]3-3¢)
(Applied Biosystems) essentially as described previously (Bailey et al., 2001) except that cells
were grown in the presence of BrdU:C (3:1 ratio, 10 M total concentration) for 18 hr and
incubated in demecolcine (0.1 g/ml) for the final 2 hr. For simultaneous visualization of both
strands, following the standard CO-FISH degradation, both FITC-TelC and TAMRA-TelG probes
were added to the hybridization buffer to 0.5 g/ml. The slides were then denatured at 80°C,
hybridized at room temperature, and processed as above.
Supplemental Figures
Supplemental Figure S1. Telomeric ChIP Monitoring the effect of TRF2B on the Presence of
Telomeric Proteins at Chromosome Ends
(A) ChIP on HeLa cells infected with pLPC or pLPC-TRF2B using the indicated antibodies.
Duplicate blots were hybridized with a TTAGGG repeats or an Alu repeat probe.
(B) Quantified telomeric ChIP data.
Supplemental Figure S2. Illustration and Examples of Methods Used for Quantification of
Telomeric Signal Loss
See Experimental Procedures for further details. The gray boxes indicate the areas in the gels
that were quantified. Normalization of human blots with a genomic DNA probe is shown in
Supplemental Figure S4.
Supplemental Figure S3. Preferential Deletion of Leading Strand Telomeric Signals Detected by
CO-FISH
(A) CO-FISH schematic. Cells are grown in the presence of BrdU and BrdC for one S phase.
After preparation of metaphase spreads, DNA is treated with UV light, which nicks DNA
preferentially at sites of halogenated pyrimidine incorporation. After treatment with ExoIII, the
newly synthesized DNA strand is degraded and only parental telomeric sequences remain. The
parental strands are then detected with strand-specific FISH.
(B) Lagging strand telomeres are largely preserved in TRF2B-expressing metaphases.
Metaphase spreads were prepared from NIH/3T3 cells 48 hr after introduction of vector or
TRF2Bvirus. The spreads were subjected to CO-FISH and stained with DAPI (red) and a Grich specific FISH probe FITC-(CCCTAA)3 (green).
(C) Leading strand telomeres are preferentially deleted after TRF2B expression. After
metaphases were treated with the CO-FISH protocol, they were stained with DAPI (blue) and a
lagging strand specific FISH probe TAMRA-(TTAGGG)3 (red).
Supplemental Figure S4. Abrogation of TRF2B-Mediated Telomere Deletion in Primary Nbs1
Mutant Fibroblasts
Primary Nbs1 mutant fibroblasts (GM07166B; Coriell) were infected with TRF2B or the vector
control and harvested on day 6 after selection to prepare DNA and protein. Absence of Nbs1
protein in Nbs1 fibroblasts and expression of TRF2B were confirmed by immunoblotting (data
not shown). Genomic DNA was digested with MboI and AluI and 3 g (quantitated by Hoechst
fluorometry) was fractionated and blotted for telomeric signals (left). The same blot was stripped
and reprobed to confirm equal loading using a genomic probe as described in the Experimental
Procedures (right). The blots were quantified and normalized using the signal in the right-hand
panel for normalization. The numbers below the lane represent the percentage change in
telomeric signal intensity relative to the vector control.
Supplemental Figure S5. TRF2B-Induced Relaxed Telomeric Circles in Mouse Cells
Detectable by 2D Gels
MEFs of the indicated genotypes were infected with TRF2B or vector control retroviruses and
harvested 5 days after selection. Genomic DNA from 1.5 × 106 cells was digested in plugs with
MboI and AluI and separated by CHEF gel in the first dimension and by standard 2D gel
conditions in the second dimension. DNAs were loaded with partially circularized I × HindIII
fragments resulting in comigration in the 2D gel. The blots were probed for the I × HindIII
fragments (right), stripped, and reprobed for telomere repeats (left). Arrows in the left bottom
panels point to circular telomeric DNA.
Supplemental Table S1. DNA Repair and Checkpoint Genes that Do Not Abrogate TRF2BInduced Telomere Signal Loss
Cell lines were infected with TRF2DB and a vector control in parallel, and the percentage
telomeric signal loss in TRF2DB-infected cells compared to the vector control cells was
determined as shown in Supplemental Figure S1 and described in the Experimental
Procedures. MEFs were immortalized by introduction of SV40LT, spontaneous immortalization,
or due to their p53-/- genotype.
Supplemental Table S2. Quantitation of TRF2B-Induced Telomere Signal Loss
The indicated cell lines were infected with vector or TRF2B retroviruses, and metaphases were
examined in parallel for telomeric signal loss by FISH (as shown in Figures 3A and 3B). The last
column indicates the number of chromosomes that have (partially) lost two telomeric signals,
one on each chromosome arm.
References
Bailey, S.M., Cornforth, M.N., Kurimasa, A., Chen, D.J., and Goodwin, E.H. (2001). Strandspecific postreplicative processing of mammalian telomeres. Science 293, 2462–2465.
Cohen, S., and Lavi, S. (1996). Induction of circles of heterogeneous sizes in carcinogentreated cells: two-dimensional gel analysis of circular DNA molecules. Mol. Cell. Biol. 16, 2002–
2014.
Gao, Y., Chaudhuri, J., Zhu, C., Davidson, L., Weaver, D.T., and Alt, F.W. (1998). A targeted
DNA-PKcs-null mutation reveals DNA-PK-independent functions for KU in V(D)J recombination.
Immunity 9, 367–376.
Karlseder, J., Smogorzewska, A., and de Lange, T. (2002). Senescence induced by altered
telomere state, not telomere loss. Science 295, 2446–2449.
Kraakman-van der Zwet, M., Overkamp, W.J., Friedl, A.A., Klein, B., Verhaegh, G.W., Jaspers,
N.G., Midro, A.T., Eckardt-Schupp, F., Lohman, P.H., and Zdzienicka, M.Z. (1999).
Immortalization and characterization of Nijmegen Breakage syndrome fibroblasts. Mutat. Res.
434, 17–27.
Loayza, D., and De Lange, T. (2003). POT1 as a terminal transducer of TRF1 telomere length
control. Nature 424, 1013–1018.
Smogorzewska, A., and de Lange, T. (2002). Different telomere damage signaling pathways in
human and mouse cells. EMBO J. 21, 4338–4348.
Takai, H., Smogorzewska, A., and de Lange, T. (2003). DNA damage foci at dysfunctional
telomeres. Curr. Biol. 13, 1549–1556.
van Steensel, B., and de Lange, T. (1997). Control of telomere length by the human telomeric
protein TRF1 Nature 385, 740–743.
van Steensel, B., Smogorzewska, A., and de Lange, T. (1998). TRF2 protects human telomeres
from end-to-end fusions. Cell 92, 401–413.
Zhu, C., Bogue, M.A., Lim, D.S., Hasty, P., and Roth, D.B. (1996). Ku86-deficient mice exhibit
severe combined immunodeficiency and defective processing of V(D)J recombination
intermediates. Cell 86, 379–389.
Zhu, X.D., Niedernhofer, L., Kuster, B., Mann, M., Hoeijmakers, J.H., and de Lange, T. (2003).
ERCC1/XPF Removes the 3¢ overhang from uncapped telomeres and represses formation of
telomeric DNA-containing double minute chromosomes. Mol. Cell 12, 1489–1498.